Consolidated the legacy audio-only identity endpoint and the multimodal identities endpoint into one multimodal API at /api/v2/identity, so voice and face identities are created and searched in one place.